The long-term goal of this project (Project 1) is to implement magnetic resonance imaging (MRI) to monitor temperature during hyperthermia treatment of the most aggressive locally advanced tumors. In this proposal the focus will be on tumors of the extremities, breast or pelvis/abdomen,. In previous work on this grant, we have shown that MRI can be effective and accurate for visualizing temperature changes in tissues associated with the absorption of nonionizing radiation. The thermal, spatial and temporal resolutions achieved with MRI vary with the therapeutic applications, but they are generally in the ranges 0.5-1.0'C, 0.5- 1.0 cm3 voxel size, and 6-20 seconds measurement acquisition time. These resolution ranges imply that MRI can play an important role as a non-invasive tool in the development and application of thermal therapies. Energy absorption rate deposition (ARD) and temperature are important metrics for gauging the overall accuracy and success of non-invasive thermometry and for implementation of control algorithms for power control (Project 2). Both of these parameters will be measured in this project. In particular, we pose the overall hypothesis that: Magnetic resonance imaging can be used to monitor ARD and temperature noninvasively throughout the tumor and surrounding normal tissue. We will test this hypothesis using the following specific aims: SPECIFIC AIM 1: Develop robust MRI thermal imaging techniques for clinical applications. SPECIFIC AIM 2: Perform 3-dimensional MRI temperature monitoring during hyperthermia treatment in patients having tumors of the lower extremity or breast. SPECIFIC AIM 3: Perform 3-dimensional MRI temperature monitoring during hyperthermia treatment in patients having tumors of the lower abdomen and pelvis Combined with the tools developed in Project 2, temperature data from MRI will, for the first time, enable 3- dimensional control of thermal dose in clinically important treatment situations.